Abstract B66: Identification of genetic and non-genetic dependencies in glioblastoma cell lines targeted by small molecules

Abstract

Abstract Introduction. Glioblastoma multiforme (GBM) is the most common and malignant brain tumor. This single histopathological entity includes a devastating collection of heterogeneous high-grade tumors showing a broad spectrum of cellular morphologies and molecular diversity. These tumors display a uniform and very short survival time. Until now and in spite of improvements achieved in neurosurgery tools and molecular biology understanding, there is still no effective treatment. For this purpose we pursue a systematic analysis of genetic and non-genetic features that correlate with chemical sensitivity in order to identify molecular dependencies. Experimental procedures. We profiled 430 small chemical compounds at 16 different duplicated concentrations in 78 GBM cell lines belonging to two different models and collections: Patient Derived GBM Cell Lines (PDGCL) and Long Term GBM Cell Lines (LTGCL) from gCCLE collection. All cell lines were deeply characterized at genotype and phenotype level. As non-genomic determinants we considered cell line model, growth rate, passages, origin, behavior, stem cell markers (OLIG2, SOX2, NESTIN), differentiation markers (GFAP, O4, TUJ1), stem cell clonality markers (CD44, CD133, CD15) and overall sensitivity displayed by all cell lines across the drug panel. As genomic determinants we take into account significant copy number variations (CNV) and driver mutations (Whole Exome Sequencing). Mutations and CNV were integrated to determine oncogene activation and tumor suppressor gene inactivation of genes involved in three major signaling pathways (p53, RB and TRK). At transcriptomic level we considered the expression of 20.647 genes and transcription patters depicted according molecular fingerprints described in high-grade gliomas (Proneural, Mesenchymal, Proliferative), and primary GBMs (Proneural, Neural, Classical, Mesenchymal). At the end 28.503 features, were surveyed against 381 filtered small chemical profiles, rendering a total 10.859.643 interactions. Summary. Three major types of chemical sensitivities associated with molecular determinants were identified: 1) Dependencies related with genotype; sensitivities specifically driven by genetic alterations (TP53 mutations and CDKN1A overexpression to MDM2 inhibitor). 2) Dependencies related with phenotype; sensitivities specifically ruled by gene expression changes associated with Mesenchymal features (VEGFR, MET, PDGFRA inhibitors), or un-specifically driven by epigenetic changes (HDCA inhibitors). 3) Dependencies dominated by enzymatic activities (NAMPT inhibitors) indirectly associated with genetic alterations or gene expression changes. Statement. Obtained results explained chemical sensitivities targeting specific genotype and molecular phenotype features (differentiation state and chromatin status determined by gene expression) represented in two models of cell lines: PDGCL and LTGCL. Both states of cellular phenotype might co-exist at GBM tumor in different proportions. Our analysis provided molecular insights to drive targeted therapies in the new era of precision medicine. Citation Format: Ruben Ferrer-Luna, Shakti H. Ramkissoon, Lori A. Ramkissoon, Rebecca Lamothe, Jaime H. Cheah, Steven E. Schumacher, Sam Haidar, Paul A. Clemons, Alykhan Shamji, Keith L. Ligon, Rameen Beroukhim. Identification of genetic and non-genetic dependencies in glioblastoma cell lines targeted by small molecules. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr B66.